Why IMAP is a good model for cloud services

I do admit it: I like the IMAP pro­to­col. I regu­lar­ly use mul­ti­ple com­pu­ters and my iPho­ne, and I read and wri­te email on all of them. IMAP makes that con­ve­ni­ent: I have the same view of my fol­ders and my inbox on every com­pu­ter. We also use a Web­mail cli­ent that uses IMAP as its under­ly­ing tech­no­lo­gy, so even via web­mail, ever­y­thing looks the same. And apart from the fact that I quite enjoy the way that I can look for all stuff that was sent my way, I much more like the fact that I have just one fol­der of email that I sent out. It does­n’t mat­ter whe­re I am when I send some­thing off: it all ends up in my sent box on our mail server. 

IMAP is a well-spe­ci­fied pro­to­col. One can argue whe­ther it’s a well-des­gi­ned pro­to­col, or whe­ther parts of it are a total night­ma­re to under­stand and imple­ment. But is is that, well-docu­men­ted. Given enough pro­gramming talent, you can sit down and wri­te eit­her a cli­ent or a ser­ver for it. (And given the track record of various IMAP cli­ents in the wild, it does take a cer­tain kind of dedi­ca­ti­on and a good load of skill to real­ly get it right.) But it’s not a tech­no­lo­gy that lets you guess what a cer­tain field on all requests might mean or why the ans­wers look so dif­fe­rent on every second fri­day of a month start­ing with J. 

The fact that it’s docu­men­ted means mul­ti­ple imple­men­ta­ti­ons exist. That means if you want, you can set up an IMAP ser­ver and just use that; or pay some­bo­dy to do just that. Per­so­nal­ly, I’m not so fond of the idea of giving all of my email away to some­bo­dy who I don’t real­ly know all so well, so my IMAP store is on a ser­ver that we run our­sel­ves. But if your pre­fe­ren­ces are dif­fe­rent, the­re are ple­nty of ser­vices that allow you to use their IMAP ser­ver, and be hap­py with that. 

This is whe­re I belie­ve cloud ser­vices should be hea­ding. Like so many, I’m a fan of Ever­no­te (I’ve writ­ten about that). I’m impres­sed by what Goog­le Docu­ments can do insi­de the brow­ser. But for eit­her com­pa­ny: do I know who else has access to my data? What laws are even appli­ca­ble for stuff that I put up? I’m sure that both Goog­le and Ever­no­te are sub­ject to US sub­poe­nas, but what about ger­man legal demands to hand over data? Or, say, tho­se ori­gi­na­ting in India? What hap­pens to all the data should Ever­no­te or Goog­le fold? I’d love Ever­no­te even more if the­re were a way to run a ser­ver of my own – becau­se then I know for sure who has access to my data. Or the pro­to­col they use were well-spe­ci­fied so that others could also con­tri­bu­te to a public ser­ver my Ever­no­te cli­ent con­nects to. 

Ah, if it only were so easy as with IMAP. 

Beitrag veröffentlicht






Schreibe einen Kommentar

Deine E-Mail-Adresse wird nicht veröffentlicht. Erforderliche Felder sind mit * markiert